1: \documentclass[12pt,preprint]{aastex}
2: \usepackage{color}
3:
4: \hyphenation{cos-mo-lo-gy}\hyphenation{cos-mo-lo-gi-cal}
5: \hyphenation{ob-ser-ving}\hyphenation{WMPol}
6:
7: \slugcomment{To appear in ApJS.} \shorttitle{The White Mountain
8: Polarimeter} \shortauthors{Levy et al.}
9:
10: \begin{document}
11:
12: \title{The White Mountain Polarimeter Telescope and an Upper Limit on CMB Polarization}
13:
14: \author{Alan R. Levy,\altaffilmark{1,2,3,4} Rodrigo Leonardi,\altaffilmark{1,5}
15: Markus Ansmann,\altaffilmark{1} Marco Bersanelli,\altaffilmark{6}
16: Jeffery Childers,\altaffilmark{1,3} Terrence D.
17: Cole,\altaffilmark{1} Ocleto D'Arcangelo,\altaffilmark{7} G. Vietor
18: Davis,\altaffilmark{1} Philip M. Lubin,\altaffilmark{1,3} Joshua
19: Marvil,\altaffilmark{1,3,8} Peter R. Meinhold,\altaffilmark{1,2,3}
20: Gerald Miller,\altaffilmark{1,9} Hugh O`Neill,\altaffilmark{1,3}
21: Fabrizio Stavola,\altaffilmark{10} Nathan C.
22: Stebor,\altaffilmark{1,2,3} Peter T. Timbie,\altaffilmark{11}
23: Maarten van der Heide,\altaffilmark{1} Fabrizio
24: Villa,\altaffilmark{10} Thyrso Villela,\altaffilmark{5} Brian D.
25: Williams,\altaffilmark{1,2,3} and Carlos A.
26: Wuensche\altaffilmark{5}}
27:
28: \email{Alan\_Levy@Raytheon.com}
29:
30: \altaffiltext{1}{Department of Physics, University of California,
31: Santa Barbara, CA 93106} \altaffiltext{2}{University of California,
32: White Mountain Research Station, Bishop, CA 93514}
33: \altaffiltext{3}{University of California, Santa Barbara Center for
34: High Altitude Astrophysics at White Mountain, University of
35: California, Santa Barbara, CA 93106} \altaffiltext{4}{Now at
36: Raytheon Vision Systems, Goleta, CA 93117}
37: \altaffiltext{5}{Instituto Nacional de Pesquisas Espaciais,
38: Divis\~{a}o de Astrof\'{i}sica, Caixa Postal 515, S\~{a}o Jos\'{e}
39: dos Campos, SP 12210-070, Brazil} \altaffiltext{6}{Dipartimento di
40: Fisica, Universit\`{a} degli Studi di Milano, Via Celoria, 16, 20133
41: Milano, Italy} \altaffiltext{7}{IFP-CNR, Via Roberto Cozzi, 53,
42: 20125 Milano, Italy} \altaffiltext{8}{Now at Department of Physics,
43: New Mexico Institute of Mining and Technology, Socorro, NM 87801}
44: \altaffiltext{9}{Now at California Institute of Technology,
45: Pasadena, CA 91125}\altaffiltext{10}{INAF-IASF Bologna, Via P.
46: Gobetti, 101, 40129 Bologna, Italy}\altaffiltext{11}{Department of
47: Physics, University of Wisconsin, Madison, WI 53706}
48:
49: \begin{abstract}
50: The White Mountain Polarimeter (WMPol) is a dedicated ground-based
51: microwave telescope and receiver system for observing polarization
52: of the Cosmic Microwave Background. WMPol is located at an altitude
53: of 3880 meters on a plateau in the White Mountains of Eastern
54: California, USA, at the Barcroft Facility of the University of
55: California White Mountain Research Station. Presented here is a
56: description of the instrument and the data collected during April
57: through October 2004. We set an upper limit on $E$-mode polarization
58: of 14 $\mu\mathrm{K}$ ($95\%$ confidence limit) in the multipole
59: range $170<\ell<240$. This result was obtained with 422 hours of
60: observations of a 3 $\mathrm{deg}^2$ sky area about the North
61: Celestial Pole, using a 42 GHz polarimeter. This
62: upper limit is consistent with $EE$ polarization predicted from a
63: standard $\Lambda$-CDM concordance model.
64: \end{abstract}
65:
66: \keywords{cosmic microwave background --- cosmology: observations
67: --- instrumentation: polarimeters}
68:
69: \section{Introduction}
70:
71: Polarization of the Cosmic Microwave Background (CMB) results from
72: Thomson scattering of CMB photons from local quadrupolar temperature
73: anisotropies during last scattering. Measurements of CMB
74: polarization promise to help to constrain cosmological models, as
75: discussed in, for example, \cite{hu97}. To date, detections of
76: $E$-mode CMB polarization have been reported by DASI
77: \citep{kovac02,leitch05}, CBI \citep{readhead04}, CAPMAP
78: \citep{barkats05}, BOOMERanG \citep{montroy06}, WMAP \citep{page07},
79: and MAXIPOL \citep{wu_jh07}.
80:
81: The White Mountain Polarimeter (WMPol) is a ground-based telescope
82: and receiver system designed to observe CMB polarization. This
83: article describes the WMPol instrument and the observing site,
84: including systems used to control WMPol remotely and monitor the
85: weather. Also presented is a summary of the data collected during
86: 2004 and the data analysis methods and results.
87:
88: \section{WMPol Instrument}
89:
90: The WMPol instrument, consisting of a microwave telescope and
91: receiver, is located at an altitude of 3880 meters in the White
92: Mountains east of Bishop, California at longitude 118\degr
93: 14\arcmin 19\arcsec W, latitude 37\degr 35\arcmin 21\arcsec N. The
94: University of California White Mountain Research Station (WMRS)
95: operates the Barcroft Facility, which includes the Barcroft
96: Observatory where the telescope is located. WMPol was installed in
97: the observatory in September 2003 and a concerted data taking
98: effort ran from April to October 2004.
99:
100: \subsection{Telescope}
101:
102: The WMPol telescope is an off-axis Gregorian telescope with a
103: 2.2-meter diameter parabolic primary. The telescope design, which
104: obeys the Dragone-Mizugutch condition \citep{Dragone78,Mizugutch76},
105: is similar to that of BEAST, a telescope dedicated to mapping CMB
106: temperature anisotropies \citep{childers05, figueiredo05,
107: meinhold05, mejia05, odwyer05, donzelli06}, and uses identical
108: aluminum coated carbon fiber reflectors. Figure~\ref{od} shows the
109: optical design including the primary, 0.9-meter diameter ellipsoidal
110: secondary, and the dewar that houses the receivers.
111:
112: The WMPol telescope is mounted on top of a table that rotates in
113: azimuth. The table is attached to a shallow cone that rests on
114: four smaller conical roller bearings. A motor drives the motion in
115: azimuth by rotating one of the roller bearings through a harmonic
116: reducing gear.\footnote{Harmonic Drive LLC, Peabody, MA 01960,
117: http://www.harmonicdrive.net/} This table, as described in
118: \cite{Meinhold93}, was used previously to observe CMB temperature
119: anisotropies from the South Pole.
120:
121: The telescope itself consists of a section that is mounted directly
122: to the table and another section (held by two bearings) that can
123: move in elevation via a linear actuator. The motors for the azimuth
124: drive and elevation actuator are controlled by a dual-axis motion
125: controller\footnote{Galil Motion Control, Rocklin, California 95765,
126: http://www.galilmc.com/} that uses relative encoders for feedback
127: and linear servo amplifiers\footnote{Western Servo Design, Carson
128: City, NV 89706, http://www.wsdi.com/} to drive the motors. Removable
129: switches that limit the motion in azimuth to approximately
130: $\pm5\degr$ were installed to prevent the possibility of damage to
131: the telescope during remote operation. Figure~\ref{te} shows the
132: major elements of the telescope.
133:
134: The position of the telescope in azimuth and elevation is measured
135: by two 16-bit absolute encoders\footnote{Gurley Precision
136: Instruments, Troy, New York 12180, http://www.gurley.com/} and tilt
137: is measured by two biaxial clinometers,\footnote{Applied
138: Geomechanics Inc., Santa Cruz, CA 95062,
139: http://www.geomechanics.com} one with a $\pm10\arcmin$ range and the
140: other with a $\pm10\degr$ range. In practice, tilt of the telescope
141: can be controlled to within $\sim\pm 2\arcmin$ over the full
142: $360\degr$ azimuth range of motion.
143:
144: A PC operates the telescope using code written at UCSB. This code
145: controls the telescope motion and reads position and tilt, takes
146: data through a data acquisition card,\footnote{Measurement
147: Computing Corp., Middleboro, MA 02346,
148: http://www.measurementcomputing.com/} operates an optical CCD
149: camera, and controls the calibrator as described in
150: Section~\ref{ssec:cal} below. Temperature at two points on each
151: reflector is measured using AD590\footnote{Analog Devices Inc.,
152: Norwood, MA 02062-9106, http://www.analog.com} ambient temperature
153: sensors.
154:
155: A great deal of effort went into arranging the power connections
156: to the telescope to suppress noise. Separate isolated linear power
157: supplies\footnote{Power-One, Camarillo, CA 93012,
158: http://www.power-one.com/} are used for the receiver and
159: housekeeping electronics and isolation transformers are used on
160: the computer and receiver power supplies. All telescope power
161: comes from a dedicated ferroresonant uninterruptible power supply,
162: or FERRUPS,\footnote{Powerware, Eaton Corp., Raleigh, NC 27615,
163: http://www.powerware.com/} and separate cables from the FERRUPS
164: bring power to the telescope drive, telescope computer, and
165: receiver. In addition, the linear servo amplifiers are installed
166: on the azimuth and elevation drives specifically to reduce noise.
167:
168: \subsection{Receiver}
169:
170: The microwave receiver system consists of two continuous comparison
171: polarimeters, one in Q-band ($38-46$ GHz) and one in W-band ($82-98$
172: GHz) plus a W-band radiometer to measure CMB temperature
173: anisotropies. Figure~\ref{fil} shows the Q-band and W-band filter
174: bands. The front ends of the receivers are contained inside an
175: evacuated vacuum vessel (dewar) and cooled to below 30 K with a
176: two-stage Gifford-McMahon cycle cryogenic
177: refrigerator\footnote{Leybold Vacuum Products Inc., Export PA 15632,
178: http://www.oerlikon.com/leyboldvacuum/} using helium as the
179: refrigerant. The receivers view the optics through a 10.8 cm
180: diameter $50\,\mu$m thick mylar vacuum window which was measured to
181: have a radiometric temperature of 3 K in W-band. A 0.3 cm thick
182: sheet of microwave transparent extruded polystyrene (``blue foam'')
183: is attached to the radiation shield inside the dewar between the
184: receivers and the window to block infrared radiation and thus reduce
185: the heat load on the cold stage.
186:
187: The receiver topology is similar to the design of the NASA WMAP
188: radiometers \citep{jarosik03} and the baseline design for the Low
189: Frequency Instrument (LFI) for ESA-Planck \citep{seiffert02}. The
190: polarimeters are designed to measure the (local - i.e. in the
191: telescope reference frame) $Q$ Stokes parameter, or $\langle
192: E^2_{hor}\rangle - \langle E^2_{ver}\rangle$, while the anisotropy
193: measuring radiometer is designed to measure the temperature
194: difference between two points on the sky separated by $\sim
195: 1^{\circ}$.
196:
197: As shown in Figure~\ref{ds}, conical corrugated scalar feed horns
198: \citep{villa97, villa98} couple the microwave radiation from the
199: telescope to the polarimeters. As is the case for other CMB
200: polarization measurements, WMPol calls for feed horns with a
201: highly symmetric radiation pattern, extremely low
202: cross-polarization, low loss over a relatively large bandwidth
203: ($\sim 20\%$ in our case), and optimal sidelobe rejection. The
204: optimized electromagnetic design for the W-band feeds required
205: throat section grooves $\sim 0.13$ mm thick and $\sim 1.5$ mm
206: deep. To meet these requirements, the W-band feeds were
207: manufactured via electroforming, with a measured deviation from
208: the theoretical mandrel dimensions of less than 10 microns. The
209: larger physical size of the Q-band feeds and their corrugations
210: made it possible to machine the horns from an aluminum cylinder
211: rather than electroforming. The main design parameters and
212: measured performances for the WMPol feeds are shown in
213: Table~\ref{t1}. Figure~\ref{rl} displays the return loss of the
214: Q-band and W-band feeds.
215:
216: For each polarimeter, the radiation from a horn feeds into an
217: orthomode transducer (OMT). The OMTs use an asymmetric design with
218: specifications as reported in Table~\ref{t2}. From the OMT, the
219: two polarizations are split and go into the two inputs of a $180$
220: degree hybrid coupler. The two outputs of the $180$ degree hybrid
221: lead to cryogenic amplification with High Electron Mobility
222: Transistor (HEMT) monolithic microwave integrated circuit (MMIC)
223: amplifiers and then more amplification outside of the dewar, plus
224: bandpass filtering in the case of the W-band polarimeter, before
225: leading to the second 180 degree hybrid. The path between the two
226: hybrid couplers is phase-sensitive and $180\degr$ phase switches
227: are installed to facilitate changing one of the path lengths by
228: $\lambda/2$ at a modulation frequency of $100 - 250$ Hz.
229: Adjustable attenuators are also installed in the phase-sensitive
230: paths to match the gain of the two paths and thin brass shims
231: (approximately $\lambda/40$ or 0.23 mm for Q-band and 0.10 mm for
232: W-band) are used to match the phase lengths of the two paths. The
233: outputs of the second hybrid go to square-law detector diodes. The
234: bandpass filters for the Q-band receiver are placed between the
235: hybrid and the detector diodes. The diode outputs are fed into a
236: $\times100$ gain differential amplifier before going to a lock-in
237: amplifier referenced to the phase modulation frequency. The
238: purpose of differencing the detector diodes is to improve the
239: 1/$f$ noise performance of the instrument. Finally, the output of
240: the lock-in goes to an ideal integrator and the data acquisition
241: card in the telescope computer. The individual diode signal levels
242: after $\times100$ gain are also recorded. These DC levels do not
243: benefit from the improved 1/$f$ performance that differencing
244: provides and do not go to a lock-in amplifier. However, these
245: measurements were found to be very helpful for calibration and
246: measuring sky temperature as described further below.
247: Table~\ref{t3} identifies the sources of various elements of the
248: polarimeters and Figure~\ref{dp} is a photograph of the inside of
249: the dewar.
250:
251: The W-band CMB temperature anisotropy radiometer is similar to the
252: polarimeters except that two horns feed into the first hybrid
253: coupler and the outputs of the second hybrid coupler feed into a
254: single pole double throw PIN switch modulated at 5 kHz which then
255: feeds a square law detector diode. Phase sensitive detection at
256: this modulation frequency provides a signal that is proportional
257: to the temperature difference between the two points on the sky
258: that the horns view. Due to space limitations and other
259: considerations, adjustable attenuators are not used in the phase
260: sensitive paths; instead, the gains are matched through small
261: changes in the bias voltages.
262:
263: All the radiometer electronics and the microwave components
264: outside of the dewar are contained in a contiguous aluminum
265: shield. To reduce noise, the dewar is electrically isolated from
266: the telescope frame except for two ground connections. One ground
267: connection goes from the dewar to the cryogenic refrigerator cold
268: head and the helium compressor through the compressed gas lines
269: and electrical cable that connect the cold head to the compressor.
270: Another ground connection from the dewar to the radiometer power
271: supply ground ensures that the dewar ground connection is not
272: severed if the cryocooler power cable is disconnected from the
273: wall.
274:
275: The noise temperatures of the Q-band and W-band polarimeters are
276: measured using manual calibrations of the non-differenced DC
277: channels. The resulting Q-band noise temperature is 127 K and the
278: W-band noise temperature is 120 K. The high noise temperature,
279: especially for the Q-band system, is mostly due to the quality of
280: amplifiers that were available to us at the time that the
281: receivers were assembled, tested, and deployed. As the measured
282: room temperature losses due to the OMT and hybrid coupler before
283: the cryogenic amplifiers are less than 0.5 dB for Q-band and 1.5
284: dB for W-band, the maximum contribution to the polarimeter noise
285: temperature due to front end losses, assuming the polarimeters are
286: at a physical temperature of 30 K, is 4 K for Q-band and 12 K for
287: W-band. The W-band anisotropy measuring radiometer failed soon
288: after the telescope was installed in the Barcroft Obsevatory and
289: repeated attempts to discover the problem have not been
290: successful. Table~\ref{t4} is a list of receiver noise
291: temperatures, effective bandwidths, and sensitivities for the Q
292: and W polarimeters.
293:
294: During the testing and commissioning of the instrument, it was
295: found that the polarized offsets were rather high ($\sim26$ K)
296: which limited the amount of gain that the polarimeter output could
297: be multiplied by before going to a data acquisition channel. The
298: offset is due to a change in DC signal out of the diodes when the
299: phase switches are in different states. We believe the origin of
300: the offset is the asymmetry between the two output arms of the
301: OMT, which is not correctable in our radiometer design. A circuit
302: was designed and successfully implemented to remove the offset
303: before sending the signal to the data acquisition system.
304:
305: \section{The Barcroft Observatory}
306:
307: Although microwave atmospheric emission is known to be not
308: significantly polarized \citep{hanany03}, the choice of a suitable
309: observing site is mandatory to optimize the data taking
310: efficiency, reduce systematic effects and avoid large baseline
311: drifts. The WMRS Barcroft Facility provides an excellent site for
312: CMB observations. The addition of a remote control system to
313: operate WMPol and weather monitoring at the Barcroft Observatory
314: improved our ability to collect data safely and efficiently.
315:
316: \subsection{Sky Temperature}
317:
318: \citet{marvil06} and a previous study of atmospheric emission at
319: Barcroft \citep{bersanelli95} show low water vapor content at the
320: Barcroft Facility, with observed limits as low as $\sim 0.2$ mm.
321: Assuming precipitable water vapor content in the range $0.2-6$ mm
322: and calibrating the atmospheric emission model of \citet{danese89}
323: with more recent data, we expect zenith atmospheric (sky)
324: temperatures at the centers of the WMPol bands in the ranges of
325: $9-12$ K at 43 GHz and $7-16$ K at 90 GHz. It should be noted that
326: the atmospheric emission spectrum in the $40-45$ GHz range,
327: dominated by O$_2$ broadened line complex, is extremely steep so
328: that the tail of the band can contribute significantly to the
329: zenith sky temperature (see Figure~\ref{fil} for the WMPol Q-band
330: filter band).
331:
332: The sky temperature is measured with sky dips. For WMPol, the sky
333: dips are scans in elevation from $32\degr$ to $48\degr$ at azimuth
334: $0\degr$, which give the zenith sky temperature after fitting to
335: the model
336: \begin{equation}
337: T_{\mathrm{antenna}} = T_{\mathrm{system}} +
338: \frac{T_{\mathrm{zenith}}}{\sin(\theta)}
339: \end{equation} where $\theta$ is the elevation angle. During the
340: observing campaign, the measured mean zenith sky temperature was
341: $(9 \pm 0.2)$ K and $(10 \pm 0.6)$ K for Q-band and W-band
342: respectively with measurements ranging from $8-11$ K for Q-band
343: and $9-15$ K for W-band. These values agree well with the
344: predicted values and measurements from \citet{marvil06},
345: \citet{childers05}, and \citet{bersanelli95}.
346:
347: \subsection{Remote Control System}
348:
349: We considered it quite advantageous to be able to operate the
350: telescope remotely. In order to facilitate remote operation and
351: data collection a number of elements were installed in the
352: observatory to enable monitoring and control of the telescope.
353:
354: The core of the remote control system is a Stargate automation
355: system.\footnote{JDS Technologies, Poway, CA 92064-6876,
356: http://www.jdstechnologies.com} The Stargate has the capability to
357: control X10 (i.e. on/off signals sent over power lines) devices
358: and on-board mechanical relays as well as read digital and analog
359: inputs. The Stargate can accept commands by phone line, internet,
360: or computer via a serial port. The X10 and mechanical relays are
361: used to turn on and off (or power cycle to reset, as needed) the
362: computers and other elements of the telescope.
363:
364: Another key element of the remote control system is the ability to
365: communicate with and operate the computers in the observatory from
366: other locations. This is achieved by using the Remote
367: Administrator (Radmin) software.\footnote{Famatech,
368: http://www.radmin.com/} Radmin allows for control of the computers
369: at the observatory from the Barcroft station and also from Santa
370: Barbara and potentially anywhere with an internet connection.
371:
372: For remote operation, it is important to be able to monitor the
373: weather so that the telescope may be protected by closing the
374: observatory shutter in case of high winds or precipitation. As
375: described in Section~\ref{ssec:weather}, a weather station and a web
376: camera (pointed towards north) are used to monitor sky and weather
377: conditions. The weather station data and camera images are archived
378: every ten minutes and downloaded to a web site at least every half
379: hour for easy monitoring. We also have the ability to view the web
380: camera live.
381:
382: The optical CCD camera\footnote{Watec America Corp., Las Vegas, NV
383: 89120} installed on the telescope to assist with pointing
384: reconstruction can be used to monitor cloud conditions at night by
385: checking whether or not Polaris is visible. The CCD camera is also
386: sensitive to clouds during the day.
387:
388: While it is critical for remote operations to maintain
389: communication with the observatory, the location of the Barcroft
390: Facility and potential for rough weather makes communication a
391: challenge. WMRS has installed two systems to connect to the
392: internet, a satellite connection\footnote{StarBand Communications
393: Inc., McLean, VA 22102, http://www.starband.com/} and a wireless
394: T1 connection through radios\footnote{Wi-LAN Inc., Calgary, AB
395: Canada T1Y 7K7, http://www.wi-lan.com/} at the observatory, White
396: Mountain peak (longitude 118\degr 15\arcmin 17\arcsec W, latitude
397: 37\degr 38\arcmin 04\arcsec N, altitude 4342 meters), and the WMRS
398: Owens Valley Lab, Bishop, CA. We also installed a cell phone base
399: station with our cell phone line connected to the Stargate system.
400: During the first campaign of data taking, lightning and other
401: extreme weather conditions caused some communications outages but,
402: fortunately, no damage to the telescope was suffered during these
403: incidents. As more experience is gained with the communications
404: infrastructure, it is expected to be more reliable.
405:
406: Another potential issue is power outages due to inclement weather.
407: All the critical elements in the observatory are on FERRUPS for
408: protection and short term power. A planned upgrade to put the
409: shutter control on a UPS and have the shutter close automatically if
410: there is a loss of power would resolve this issue.
411:
412: \subsection{Weather Monitoring}\label{ssec:weather}
413:
414: To help protect the WMPol telescope from inclement weather and in
415: order to aid data selection, a commercially available weather
416: station\footnote{Davis Instruments Corp., Hayward, CA 94545,
417: http://www.davisnet.com/} and USB web camera\footnote{Creative Labs
418: Inc., Milpitas, CA 95035, http://us.creative.com/} were installed at
419: the Barcroft Observatory. The weather station has the ability to
420: measure temperature, relative humidity, barometric pressure, wind
421: speed and direction, and solar irradiance. Relative humidity in
422: conjunction with the web camera during the day and a CCD camera used
423: for imaging of stars during the night are used to determine the
424: onset of bad weather. To avoid damage to the telescope, the shutter
425: was closed and data taking suspended if inclement weather was
426: imminent or if the average wind speed exceeded 20 mph (8.9
427: m$\cdot\mathrm{s}^{-1}$) and/or wind gusts exceeded 30 mph (13
428: m$\cdot\mathrm{s}^{-1}$).
429:
430: During the April through October 2004 observing campaign, the mean
431: temperature was $3.3\degr$C with an average daily temperature swing
432: of approximately $8\degr$C and the wind speed was lower than our
433: damage-avoidance threshold 93 percent of the time. \citet{marvil06}
434: give a more detailed analysis of the Barcroft Observatory weather
435: station data
436:
437: The web camera images were obtained every 10 minutes and were rated
438: by eye. During the observing campaign, the sky was ``Clear,''
439: corresponding to a cloudless period, $46\%$ of the time and ``Partly
440: Cloudy,'' indicating some scattered clouds, $30\%$ of the time.
441: Periods that were rated ``Mostly Cloudy'' ($15\%$ of the time) or
442: ``Overcast'' ($9\%$ of the time) were considered to be times when
443: the weather conditions would not result in useful radiometer data.
444:
445: \section{Observing Campaign}
446:
447: After the telescope was installed in the Barcroft Observatory,
448: Winter 2003-04 was spent commissioning and testing the telescope. We
449: define 2004 April 23 as the official beginning of the observing
450: campaign which ended on 2004 October 17 with the first winter storm
451: of the year and the scheduled seasonal closure of the Barcroft
452: Facility.
453:
454: \subsection{Data Acquisition}
455:
456: During the 178-day campaign, data were recorded during 2169 hours,
457: or just over half of the available time. Table~\ref{t5} describes
458: the observing time lost due to various issues. Inclement weather,
459: including precipitation and high wind speed, was a major factor
460: and the Stargate automation system was damaged on multiple
461: occasions, we believe due to lightning, causing a great deal of
462: down time. Several levels of surge protection were added to the
463: Stargate during the observing campaign but the seasonal lightning
464: storms ended before it could be determined whether this solved the
465: problem.
466:
467: The cryogenic refrigerator used to cool the front ends of the
468: radiometers also halted data collection for two reasons. First,
469: the cold end of the refrigerator is known to have a mechanical
470: problem that slightly reduces performance and seems to cause the
471: cooler and dewar to warm up on its own after about fourteen days
472: of continuous operation. In order to recycle the system, the
473: inside dewar temperature has to be brought up all the way to
474: ambient temperature ($\gtrsim 280$ K) before cooling down again;
475: this process requires approximately 24 hours. The other reason for
476: data loss was overheating of the refrigerator's helium compressor
477: during the middle of the day during summer. The compressor is
478: liquid-cooled with a fan plus radiator to draw heat away from the
479: compressor motor. A thermal cutout prevents damage due to
480: overheating. During winter, a $50-50$ water-antifreeze (ethylene
481: glycol) mix is used to prevent the coolant from freezing during
482: times when the compressor is shut down. This mix has reduced
483: thermal properties as compared to pure water and is not sufficient
484: to transfer heat during midday from the compressor motor in the
485: increased summer temperature conditions. Once this issue was
486: discovered, replacing the anti-freeze mix with distilled water
487: solved the problem.
488:
489: \subsection{Pointing Reconstruction}
490:
491: In order to determine our pointing, we used the Moon and stars to
492: align the optical CCD camera (approximate field of view of $2\fdg7
493: \times 2\fdg0$) with the radiometric channels. The size of each CCD
494: camera image pixel, $15\arcsec \pm 1\arcsec$, was determined by
495: observing known stars fields. Observations of the Moon were used to
496: find the central pixel positions of the radiometers in the CCD
497: camera. Subsequent observations of the pixel position of Polaris in
498: the CCD camera allowed us to calculate any offset between the
499: polarimeters and the position of the telescope as defined by the
500: encoders. We estimate our pointing error to be $\pm3\arcmin$ for the
501: Q-band polarimeter and $\pm2\arcmin$ for the W-band polarimeter.
502: This pointing error is a small fraction of the beam size. To verify
503: that there were no changes in the pointing on a day-to-day basis,
504: CCD camera images were recorded at all times at a rate around one
505: per minute. In addition we spent 1200 seconds each night of
506: observation staring at Polaris and $\lambda$ Ursae Minoris at
507: approximately 7:00 UT to build up a data base to reveal any changes
508: in the pointing of the telescope or CCD camera.
509:
510: \subsection{Observing Strategy}
511:
512: WMPol used a constant elevation scan strategy, similar to the
513: strategies employed by PIQUE \citep{hedman01, hedman02}, COMPASS
514: \citep{farese03, farese04}, and CAPMAP \citep{barkats05}. The
515: telescope maintained an elevation of $37\fdg6$ while scanning the
516: sky in azimuth $\pm45\arcmin$ about the North Celestial Pole (NCP)
517: with a period of 9.5 seconds. An advantage of this strategy is
518: that the atmospheric contribution remains nearly constant over
519: short time intervals. Sky rotation allows the telescope to observe
520: a circular region on the sky centered on the NCP. The useful
521: region of the sky observed has an area of 5.5 $\mathrm{deg}^2$ in
522: the Q-band polarimeter channel and 1.1 $\mathrm{deg}^2$ for the
523: W-band polarimeter channel. The larger coverage area of the Q-band
524: polarimeter is due to a $54\arcmin \pm 1\arcmin$ offset of the
525: Q-band beam from the optical axis, the value of which was
526: determined by observations of the Moon with both polarimeters.
527:
528: WMPol only measures the Stokes parameter $Q'$ in the frame defined
529: by the orthomode transducer of the instrument. The sky Stokes
530: parameters $Q$ and $U$, expressed according to IAU convention
531: \citep{iau74}, are related to WMPol measurements by
532: \begin{equation}\label{e:QUtransform}
533: Q'\propto Q\cos2\psi+U\sin2\psi,
534: \end{equation}
535: where $\psi$ is the parallactic angle measured from North and
536: increasing through East (i.e. the angle at the intersection between
537: two great circles passing through the observation point, one
538: containing the NCP and the other the WMPol zenith). Due to the WMPol
539: scan strategy, $\psi$ remains constant and approximately equal zero
540: during data acquisition. So, in the absence of noise, Equation
541: (\ref{e:QUtransform}) allows us to interpret $Q'\simeq Q$. Due to
542: the choice of this scan strategy, WMPol measurements of $U$ Stokes
543: parameter are negligible and are not considered during further
544: analysis.
545:
546: \subsection{Beam Determination}
547:
548: The beam sizes of the radiometers on the sky after passing through
549: the optics were modeled using GRASP8.\footnote{TICRA, Copenhagan,
550: Denmark, http://www.ticra.com} Geometrical optics with geometrical
551: theory of diffraction was used for the secondary reflector, while
552: for the primary reflector physical optics was applied. The
553: geometry was analyzed in transmitting mode (i.e. from feed to
554: sky). In the calculations, far field effects were neglected at the
555: feed level as the feed aperture is only about $4.5\lambda$ and the
556: secondary reflector is at the far field of the feeds. Gaussian
557: beam models of the feeds for both Q-band and W-band were used with
558: modeled FWHM beam sizes of $19.4$ degrees at 41.5 GHz and $19.3$
559: degrees at 92.5 GHz. Each beam was calculated using a regular
560: (u,v) grid centered on each beam peak in the range (-0.02, 0.02)
561: at 41.5 GHz and (-0.01, 0.01) at 92.5 GHz, for both u and v,
562: corresponding to $1\fdg15\times1\fdg15$ and $0\fdg575\times
563: 0\fdg575$ on the sky. Figure~\ref{bmQP} shows contour plots of the
564: beam models for the polarimeters. To characterize each beam, an
565: elliptical Gaussian fit of the beam in the whole angular region of
566: calculation was obtained. The modeled characteristics are reported
567: in Table~\ref{t6} with expected beam sizes (average FWHM) of the
568: Q-band and W-band polarimeters of $18\farcm8$ and $8\farcm5$
569: respectively.
570:
571: Measurements of the actual beam size were accomplished using the
572: Moon and Tau A. A least squares fit between the observations and the
573: source convolved with a Gaussian beam was used to determine the beam
574: size. The measured FWHM beam size of the Q-band polarimeter is
575: $24\arcmin \pm 3\arcmin$ based on observations of Tau A, as shown in
576: Figure \ref{f:beamsize_Q_chi_reduced.eps}. This result is similar to
577: the $22\arcmin \pm 2\arcmin$ quoted in \citet{childers05}.
578: Presumedly due to 1/$f$ issues, Tau~A was not detected with the
579: W-band channel but, by using Moon observations and a model for the
580: Moon as given by \citet{keihm84}, a FWHM beam size of $12\arcmin \pm
581: 3\arcmin$ was found.
582:
583: \subsection{Calibration}\label{ssec:cal}
584:
585: Calibration of the receiver was accomplished with multiple
586: techniques. A manual calibration of the non-differenced channels
587: of the polarimeters was performed by recording the radiometer
588: signals while an ECCOSORB\footnote{Emerson \& Cuming Microwave
589: Products, Randolph, MA 02368, http://www.eccosorb.com} ambient
590: temperature load was placed in front of the radiometers and
591: comparing to the signal levels when viewing the sky.
592:
593: To calibrate the polarization channels, a grid of 150 $\mu$m
594: diameter copper wires with a 485 $\mu$m spacing was used. This
595: grid was held at an angle of 45 degrees with respect to the dewar
596: window and provided a large polarized signal due to the difference
597: between the temperature of the sky and an ECCOSORB ambient
598: temperature load. As this calibration procedure was manual and
599: required changing the lock-in gain, we also implemented an
600: automated calibrator with a thin dielectric (polypropylene) sheet
601: to perform a relative calibration every 10 minutes. This technique
602: is similar to that described in detail in \citet{odell02}. For
603: each manual calibration of the polarized channels, a transfer
604: standard was calculated and used to convert from the automatic
605: calibrator. This transfer standard, $\alpha$, takes the form
606: \begin{equation}\label{e:transfer_standard}
607: \alpha = \frac{\vert \Delta V_{\mathrm{film}} \vert}{G \vert
608: T_{\mathrm{ref}} - T_{\mathrm{sky}} \vert}
609: \end{equation} where $\Delta V_{\mathrm{film}}$ is the polarimeter signal
610: offset due to placing the film in front of the dewar,
611: $T_{\mathrm{ref}}$ is the antenna temperature of the ambient
612: reference load, $T_{\mathrm{sky}}$ is the measured mean radiometric
613: sky temperature at elevation $37\fdg6$, and $G$ is the calibration
614: constant (volts per kelvin) from a manual grid calibration.
615: Figure~\ref{auto} shows examples of the signal level change in the
616: polarization channels due to the automatic calibration and
617: Figure~\ref{cal} shows the orientation of the automatic calibrator
618: (or copper wire grid for manual calibration) with respect to the
619: dewar. We estimate the calibration uncertainty on the adopted
620: procedure to be $20\%$.
621:
622: We verified our Q-band calibration by using Tau~A as a target. This
623: source was observed through 7 constant elevation scans during 41
624: minutes on 2004 September 1. The scans were combined to obtain a
625: Stokes Q map of Tau~A at a parallactic angle of $58\fdg3 \pm
626: 0\fdg4$. The map obtained was compared with Tau~A on Q-band maps
627: from the WMAP sky survey \citep{page07, hinshaw07}. This procedure
628: allowed us to verify the Q-band automatic calibration within
629: measurement error.
630:
631: \section{Data Analysis}
632:
633: Scientific and housekeeping data were stored in files corresponding
634: to 20 minutes in length. For each file and channel we computed
635: statistical estimators, polarimeter white-noise levels, the weather
636: condition, calibration constants, and other information useful in
637: classifying the data. This procedure allowed us to investigate long
638: time scale behavior of instrument and data, and flag data before
639: producing Stokes Q sky maps.
640:
641: \subsection{Data Selection}
642:
643: Table~\ref{t7} gives the number of hours of data that were cut from
644: the raw data to create a CMB data set. Data files for times when the
645: telescope was not scanning about the NCP were not used except where
646: data were specifically taken for the purposes of pointing
647: reconstruction, beam size determination, and calibration as
648: discussed below. Also, files shorter than 60 seconds were cut as
649: these files were not expected to contain enough data to be useful.
650: During midday in the late spring and summer the Sun was high enough
651: in the sky to shine directly on the telescope. This not only caused
652: the secondary reflector to heat up by as much as 20\degr C but also
653: caused the warm components of the polarimeters to heat up to a
654: temperature higher than the servo set point. Changes in signal and
655: gain due to this heating caused the DC signal levels of the
656: polarization channels to exceed the limits of the data acquisition
657: electronics and we labeled these data as ``saturated.''
658:
659: The Q-band and W-band polarization-sensitive channels of each data
660: file were visually inspected. The time-ordered data before and
661: after removing the DC offset on these channels for each half scan
662: (one sweep in azimuth from $-45\arcmin$ to $+45\arcmin$ or the
663: reverse) was examined along with the power spectral density of
664: both channels. The rms of the data after removing the offset was
665: also calculated. Many files that were found to have a rms
666: significantly different from the mean had unusual features in both
667: polarization channels that were presumed to be due to clouds
668: passing through the beam. These files were removed from the data
669: analysis.
670:
671: After the rms cut, we found that there were data that passed the
672: cut that were collected during bad weather (i.e. web camera image
673: rating equal to ``Mostly Cloudy'' or ``Overcast'') and these data
674: were discarded. Another small amount of data was cut due a cold
675: plate temperature that was too high as a result of the dewar
676: warming up. After cuts, 1574 hours of raw Q-band data and 1208
677: hours of raw W-band data remained for scientific analysis. These
678: cuts rejected about 27$\%$ of Q-band data and 44$\%$ of W-band
679: data.
680:
681: \subsection{Data Reduction}
682:
683: In order to go from raw time ordered data to calibrated data, the
684: average offset per half scan, auto-calibration sequences, and time
685: domain spikes were removed, the data were calibrated, and binned
686: in azimuth.
687:
688: Although the signal rms per azimuth bin was consistent with, but
689: slightly higher than, what the instrument sensitivity would predict,
690: calibrated data binned in azimuth did reveal the presence of a scan
691: synchronous signal (SSS). The SSS was sensitive to atmospheric
692: temperature changes and was likely due to thermal emission from the
693: dome into the beam sidelobes of the telescope. A linear regression
694: of calibrated signal as a function of azimuth, performed for each 20
695: minutes observing period, revealed an average slope of 1 mK/degree
696: in both bands, with variations up to 5 mK/degree/day and 14
697: mK/degree/day in Q and W bands respectively. Several strategies were
698: tested to minimize SSS contribution, but no reliable model could be
699: established to describe the SSS as a function of ambient
700: temperature, time, or housekeeping data. The SSS removal technique
701: employed involved subtracting from each file a second-order
702: polynomial fit to the calibrated data binned in azimuth. After this
703: procedure, pointing reconstruction was performed and the resulting
704: data was submitted to a maximum likelihood map making algorithm.
705:
706: \subsection{Map Making}
707:
708: We adopted the HEALPix\footnote{http://healpix.jpl.nasa.gov/}
709: pixelization scheme \citep{gorski05}, and used the
710: MADCAP\footnote{http://crd.lbl.gov/$\sim$borrill/cmb/madcap/}
711: package to produce Stokes Q maps \citep{borrill99}. The required
712: inverse time-time noise correlation matrix in the frequency domain
713: was computed as described in \citet{stompor02}. Noise maps were
714: estimated by computing the difference between two maps, each one
715: produced with half of the available data. Finally, after removing
716: any residual offset by imposing the condition that the map has zero
717: mean, a $\chi^{2}$ test for statistical significance was performed
718: on all the sky and noise maps to check if they were consistent with
719: the null hypothesis, i.e. consistent with no signal.
720:
721: For the Q-band maps, the null hypothesis was rejected in all the sky
722: and noise maps produced when including data acquired during the day;
723: these maps contain non-stationary residuals from spurious signals
724: due to SSS and the Sun. We applied several masks to select
725: sub-regions in the maps that may be less affected by SSS but the
726: null hypothesis was also rejected in all these sub-regions. For maps
727: obtained using only data acquired during the night, the null
728: hypothesis was accepted in all the sky and noise maps produced. The
729: same result was also obtained when applying masks to select
730: sub-regions in these maps.
731:
732: For the W-band maps, the situation was different. The W-band data
733: suffered from much higher SSS contamination, and we were unable to
734: determine a filter process that could retain sky signal and remove
735: the SSS. Because of this, these maps were rejected during further
736: analysis.
737:
738: As a result, the Q-band night time data is the best data set
739: available from WMPol. The Stokes Q map obtained with this subset
740: contains 422 hours of integration, and covers an area of 3
741: $\mathrm{deg}^{2}$ in a region near the NCP with an effective
742: resolution of $24\arcmin$ FWHM at 42 GHz. The Stokes Q map contains
743: 231 pixels (HEALPix Nside=512, which corresponds to a 6.9 arcmin
744: pixel size), an error of 45 $\mu\mathrm{K}/\mathrm{pixel}$, an rms
745: of 76 $\mu\mathrm{K}$, and thus an error in the mean of 5
746: $\mu\mathrm{K}$. These 422 hours of data were collected between 2004
747: July 6 and October 17 and comprise $17\%$ of the possible observing
748: time during this 104 day period. The day time data that were
749: collected during this time period and passed all cuts but were
750: ultimately not included in the final map comprise 378 additional
751: hours.
752:
753: \subsection{Angular Power Spectra}
754:
755: From Monte Carlo simulations, the WMPol \emph{EE} transfer
756: function was computed assuming a Gaussian beam, the scan strategy,
757: the pixelization scheme, the implemented data reduction pipeline,
758: and the data analysis package. The transfer function is the ratio
759: between the output pseudo angular power spectrum and the
760: corresponding input angular power spectrum. This procedure allowed
761: us to compute the transfer function in the multipole range
762: $100<\ell<500$. The WMPol transfer function is noise dominated for
763: $\ell<120$, it becomes negligible for $\ell>300$, and it takes
764: into account that we have observed only the $Q$ Stokes parameter.
765: Hence WMPol is sensitive to the multipole range
766: $170\lesssim\ell\lesssim240$. We estimated $EE$ using the PolSpice
767: package \citep{chon04,szapudi01}, which is a method for estimating
768: angular power spectrum from the two-point correlation function and
769: deals with the effects of limited sky, beam smoothing, noise
770: contamination, and inhomogeneous pixel errors. We obtained an
771: upper limit of 14 $\mu\mathrm{K}$ with 95$\%$ confidence level.
772: This upper limit does not include the uncertainties in the Q-band
773: calibration ($20\%$), beam size ($13\%$, which corresponds to
774: $26\%$ uncertainty in beam solid angle), or beam shape, which was
775: assumed to be Gaussian. The Table~\ref{t8} compares our result to
776: recent CMB polarization measurements, Figure~\ref{windowfunction}
777: displays the WMPol transfer function and Figure~\ref{upperlimit}
778: shows our result in the context of the $\Lambda$-CDM concordance
779: model.
780:
781: \section{Conclusion}
782:
783: WMPol was designed, built, and tested at University of California,
784: Santa Barbara and installed in the WMRS Barcroft Observatory to
785: measure CMB polarization in Q-band and W-band. With this telescope
786: and receiver system 2169 hours of data were collected during 2004.
787: The WMPol telescope proved to be robust and a novel remote control
788: system using relatively inexpensive and commercially-available
789: hardware and software was successfully implemented and utilized to
790: operate the telescope over the internet.
791:
792: WMPol sets an upper limit on $E$-mode polarization of 14
793: $\mu\mathrm{K}$ ($95\%$ confidence limit, not including calibration
794: and beam uncertainties) in the multipole range $170<\ell<240$. This
795: result was obtained from 422 hours of observations of a Q-band
796: continuous comparison polarimeter on a 3 $\mathrm{deg}^2$ sky area
797: near the North Celestial Pole. The reported upper limit is based on
798: a method that estimates the CMB polarization power spectra via the
799: two-point correlation function, and it is consistent with $EE$
800: polarization predicted from a standard $\Lambda$-CDM concordance
801: model.
802:
803: \acknowledgments
804:
805: This work would not have been possible without the support of Frank
806: Powell, the Director of the White Mountain Research Station, and the
807: dedicated WMRS staff including Associate Director John Smiley, Paul
808: Addison, Dori Cann, Martin Freeman, Ryan Kitts, Matt Lokken, Richard
809: McDade, Gary Milender, DonnaLisa Shinn, Valerie Thorp, Mike Virgin,
810: and Denise Waterbury. We are also grateful for the many UCSB
811: undergraduate students who contributed during the course of
812: building, testing, and installing the telescope. The superb UCSB
813: Machine Shop was instrumental in the success of this project. We
814: thank Todd Gaier and Sandy Weinreb for providing the HEMT
815: amplifiers. We also acknowledge the efforts of Chris O'Dell who did
816: much of the initial work in designing the W-band polarimeter and
817: Miikka Kangas who prepared some of the figures in this article. We
818: thank Julian Borrill, Radek Stompor, and Chris Cantalupo for help
819: with and providing resources to use MADCAP and we thank Eric Hivon
820: and Ian O'Dwyer for help with and providing resources to use
821: PolSpice.
822:
823: This work was supported by NSF grants AST-9802851, 9813920, and
824: 0118297 and the UCSB Office of Research. The development of the
825: carbon fiber reflectors was supported by NASA grants NAG5-4078,
826: NAG5-9073, and NAG5-4185. We are especially grateful to Thin Film
827: Technology, Inc., Buellton, CA for coating the reflectors. The
828: W-band feeds were developed at the the Microwave Laboratory of
829: Istituto di Fisica dei Plasmi - CNR, Milan, Italy.
830:
831: A.L. was supported by a California Space Grant Consortium Graduate
832: Research Fellowship and A.L., N.S., and B.W. were supported by the
833: WMRS Graduate Student Research Fellowship. R.L. was supported by
834: CAPES and the NASA Planck project under JPL contract \#1261740.
835: T.V. and C. A. W. were partially supported by FAPESP grant
836: 00/06770-2. In addition, T.V. acknowledges support from CNPq grant
837: 305219/2004-9 and C. A. W. acknowledges support from CNPq grant
838: 307433/2004-8-FA and FAPESP grant 96/06501-4. G. M. was supported
839: by a UCSB Chancellor's Undergraduate Research Award and M. vdH.
840: was supported by a UCSB Undergraduate Research and Creative
841: Activities Award.
842:
843: Some of the results in this paper have been derived using the
844: HEALPix package \citep{gorski05}. The research described in this
845: paper was performed in part at the Jet Propulsion Laboratory,
846: California Institute of Technology, under a contract with the
847: National Aeronautics and Space Administration.
848:
849: \begin{thebibliography}{}
850:
851: \bibitem[Barkats et al.(2005)]{barkats05} Barkats, D., et al.
852: 2005, \apjl, 619, L127
853: \bibitem[Bersanelli et al.(1995)]{bersanelli95} Bersanelli, M., et al.
854: 1995, \apj, 448, 8
855: \bibitem[Borrill(1999)]{borrill99} Borrill, J. 1999, preprint
856: (astro-ph/9911389)
857: \bibitem[Cartwright et al.(2005)]{cartwright05} Cartwright, J. K., et al.
858: 2005, Physical Review D, 17, 1901
859: \bibitem[Childers et al.(2005)]{childers05} Childers, J., et al.
860: 2005, \apjs, 158, 124
861: \bibitem[Chon et al.(2004)]{chon04} Chon, G., et al.
862: 2004, MNRAS, 350, 914
863: \bibitem[Donzelli et al.(2006)]{donzelli06}Donzelli, S., et al. 2006, \emph{CMB and Physics of the Early
864: Universe} Proceedings, 33
865: \bibitem[Dragone(1978)]{Dragone78}Dragone, C. 1978, The Bell
866: System Technical Journal, 57, 2663
867: \bibitem[Danese \& Partridge(1989)]{danese89}Danese, L. \& Partridge, R. B. 1989, \apj, 342
868: 604
869: \bibitem[Farese et al.(2003)]{farese03} Farese, P. C., et al.
870: 2003, New Astronomy Reviews, 47, 1033
871: \bibitem[Farese et al.(2004)]{farese04} Farese, P. C., et al.
872: 2004, \apj, 610, 625
873: \bibitem[Figueiredo et al.(2005)]{figueiredo05} Figueiredo, N., et
874: al. 2005, \apjs, 158, 118
875: \bibitem[G\'{o}rski et al.(2005)]{gorski05} G\'{o}rski, K. M., et al. 2005,
876: \apj, 622, 759
877: \bibitem[Hanany \& Rosenkranz (2003)]{hanany03} Hanany, S. \& Rosenkranz, P. 2003, New Astronomy Reviews, 47, 1159
878: \bibitem[Hedman et al.(2001)]{hedman01} Hedman, M. M., et al.
879: 2001, \apjl, 548, L111
880: \bibitem[Hedman et al.(2002)]{hedman02} Hedman, M. M., et al.
881: 2002, \apjl, 573, L73
882: \bibitem[Hinshaw et al.(2007)]{hinshaw07} Hinshaw, G., et al. 2007, ApJS,
883: 170, 288
884: \bibitem[Hu \& White(1997)]{hu97} Hu, W., \& White, M. 1997, New Astronomy, 2, 323
885: \bibitem[IAU (1974)]{iau74} IAU, 1974, Transactions of the
886: IAU Vol. XVB, 166
887: \bibitem[Keating et al.(2001)]{keating01} Keating, B. G., et al. 2001, \apjl,
888: 560, L1
889: \bibitem[Keihm (1984)]{keihm84} Keihm, S. J.
890: 1984, Icarus, 60, 568
891: \bibitem[Kovac et al.(2002)]{kovac02} Kovac, J. M., Leitch, E. M., Pryke, C., Carlstrom, J. E., Halverson, N. W., \& Holzapfel W. L. 2002, Nature,
892: 420, 772
893: \bibitem[Jarosik et al.(2003)]{jarosik03} Jarosik, N., et al. 2003, \apjs, 145, 413
894: \bibitem[Leitch et al.(2005)]{leitch05} Leitch, E. M., Kovac, J. M., Halverson, N. W., Carlstrom, J. E., Pryke, C. \& Smith, M. W. E.
895: 2005, \apj, 624, 10
896: \bibitem[Marvil et al.(2006)]{marvil06} Marvil, J., et al. 2006,
897: New Astronomy, 11, 218
898: \bibitem[Meinhold et al.(1993)]{Meinhold93} Meinhold, P. R., et al. 1993,
899: \apj, 406, 12
900: \bibitem[Meinhold et al.(2005)]{meinhold05} Meinhold, P. R., et al. 2005, ApJS, 158, 101
901: \bibitem[Mej\'{i}a et al.(2005)]{mejia05} Mej\'{i}a, J., et al. 2005, ApJS, 158, 109
902: \bibitem[Mizugutch, Akagawa, \& Yokoi(1976)]{Mizugutch76} Mizugutch, Y., Akagawa, M., \& Yokoi, H. 1976, Digest of
903: 1976 AP-S International Symposium on Antennas and Propagation, 2
904: \bibitem[Montroy et al.(2006)]{montroy06} Montroy, T. E., et al.
905: 2006, \apj, 647, 813
906: \bibitem[O'Dell et al.(2002)]{odell02} O'Dell, C. W., Swetz, D. S., \& Timbie, P. T. 2002, IEEE-MTT, 50, 9, 2135
907: \bibitem[O'Dwyer et al.(2005)]{odwyer05} O'Dwyer, I. J., et al. 2005, ApJS, 158, 93
908: \bibitem[Page et al.(2007)]{page07} Page, L., et al. 2007, ApJS,
909: 170, 335
910: \bibitem[Readhead et al.(2004)]{readhead04} Readhead, A. C. S., et al.
911: 2004, Science, 306, 836
912: \bibitem[Seiffert et al.(2002)]{seiffert02} Seiffert, M., et al.
913: 2002, Astronomy and Astrophysics, 391, 1185
914: \bibitem[Stompor et al.(2002)]{stompor02} Stompor, R., et al. 2002,
915: Physical Review D, 65, 022003
916: \bibitem[Szapudi(2001)]{szapudi01} Szapudi, I., et al.
917: 2001, \apj, 548, L115
918: \bibitem[Villa, Bersanelli, \& Mandolesi(1997)]{villa97} Villa, F., Bersanelli, M.,
919: \& Mandolesi, N. 1997, ITESRE/CNR Internal Report, 188
920: \bibitem[Villa, Bersanelli, \& Mandolesi(1998)]{villa98} Villa, F., Bersanelli, M.,
921: \& Mandolesi, N. 1998, ITESRE/CNR Internal Report, 206
922: \bibitem[Wu et al.(2007)]{wu_jh07} Wu, J. H. P., et al. 2006 ApJ,
923: 665, 55
924: \end{thebibliography}
925:
926: \clearpage
927:
928: \begin{deluxetable}{lrr}
929: \tablewidth{0pt}
930: \tablecaption{Design of the WMPol Corrugated Scalar Feed Horns \label{t1}} \tablehead{\colhead{Parameter} &
931: \colhead{Q-band} & \colhead{W-band}} \startdata
932:
933: Flare angle (degrees) & 7 & 7 \\
934: Aperture diameter (mm) & 27.16 & 12.08 \\
935: Aperture corrugation depth (mm) & 1.9 & 0.88 \\
936: Aperture corrugation width (mm) & 1.45 & 0.67 \\
937: Throat diameter (mm) & 9.04 & 2.972 \\
938: Throat corrugation depth (mm) & 3.55 & 1.51 \\
939: Throat corrugation width (mm) & 1.45 & 0.13 \\
940: Number of corrugations & 34 & 37 \\
941: Corrugation step (mm) & 2.17 & 1.00 \\
942: Design return loss (dB) & $< -35$ & $< -35$ \\
943: FWHM (degrees) & 20 & 20 \\
944: Side lobe level(dB) & $< -30$ & $< -30$ \\
945:
946: \enddata
947: \end{deluxetable}
948:
949: \begin{deluxetable}{lrr}
950: \tablewidth{0pt}
951: \tablecaption{Specifications of the Orthomode Transducers \label{t2}} \tablehead{\colhead{Parameter} &
952: \colhead{Q-band} & \colhead{W-band}} \startdata
953:
954: Passband (GHz)& $36.0-43.0$\tablenotemark{*} & $85.0-105.0$ \\
955: VSWR & $1.2:1$ Max & $1.5:1$ Max \\
956: Insertion loss (dB) & $0.1$ Max & Not specified \\
957: Cross polarization (dB) & $-35$ & $-30$ \\
958:
959: \enddata
960: \tablenotetext{*}{useful band of Q-band OMT was measured to extend
961: to 46 GHz}
962:
963: \end{deluxetable}
964:
965: \begin{deluxetable}{lll}
966: \tablewidth{0pt}
967: \tablecaption{Major Elements of the Polarimeters \label{t3}} \tablehead{\colhead{Device} &
968: \colhead{Source (Q-band)} & \colhead{Source (W-band)}}
969: \startdata Corrugated scalar feed horns & Univ. of Milan (Italy) & INAF-IASF (Italy) \\
970: Orthomode transducers & Custom Microwave\tablenotemark{a} & Vertex RSI\tablenotemark{b} \\
971: Hybrid couplers & QuinStar Technology\tablenotemark{c} & Millitech\tablenotemark{d} \\
972: Cryogenic amplifiers & JPL/UCSB & JPL \\
973: Warm amplifiers & Avantek & JPL \\
974: Phase switches & Pacific Millimeter Products\tablenotemark{e} &
975: Pacific Millimeter Products \\
976: Adjustable attenuators & Aerowave\tablenotemark{f} & Aerowave\\
977: Bandpass filters & UCSB & UCSB \\
978: Microwave detector diodes & Pacific Millimeter Products & Pacific
979: Millimeter Products\\
980: Differential amplifier & UCSB & UCSB\\
981:
982: \enddata
983: \tablenotetext{a}{Custom Microwave, Inc., Longmont, CO 80501,
984: http://www.custommicrowave.com} \tablenotetext{b}{Vertex RSI,
985: Torrance, California 90505, http://www.tripointglobal.com}
986: \tablenotetext{c}{QuinStar Technology, Inc., Torrance, California
987: 90505, http://www.quinstar.com} \tablenotetext{d}{Millitech, Inc.,
988: Northampton, MA 01060, http://www.millitech.com}
989: \tablenotetext{e}{Pacific Millimeter Products, Golden, CO 80401,
990: http://www.pacificmillimeter.com} \tablenotetext{f}{Aerowave, Inc.,
991: Medford, MA, 02155, http://www.aerowave.net/}
992:
993: \end{deluxetable}
994:
995: \begin{deluxetable}{ccrc}
996: \tablewidth{0pt}
997: \tablecaption{Parameters of the Polarimeters \label{t4}} \tablehead{ \colhead{} &
998: \colhead{$T_{system}$} & \colhead{$\Delta\nu_{eff}$} &
999: \colhead{Sensitivity}\\\colhead{Channel} & \colhead{(K)} &
1000: \colhead{(GHz)} & \colhead{(mK$\cdot\mathrm{s}^{1/2}$)}}
1001: \startdata Q-band (38-46 GHz) & 127 & 8.0 & 3.5 \\ W-band (82-98 GHz) & 120 & 12.6 & 3.0 \\
1002:
1003: \enddata
1004: \end{deluxetable}
1005:
1006: \begin{deluxetable}{lr}
1007: \tablecaption{Primary Reasons for Not Taking Data \label{t5}}
1008: \tablehead{\colhead{Issue} & \colhead{Hours}} \startdata \\Weather\tablenotemark{*} \space (snow, hail, rain, high wind) & 615 \\
1009: Problems with remote control system (often caused by lightning) & 955 \\ Scheduled power outages & 71 \\
1010: Cryocooler recycling & 147 \\ Various glitches (computer/code
1011: crashes, cryocooler overheating, etc.) & 121 \\ Telescope
1012: maintenance and manual calibrations & 194\\ \\ \tableline\\
1013: Total lost observing time & 2103\\
1014: Total successful observing time (Raw data) & 2169\\
1015: Total possible observing time & 4272
1016: \enddata \tablenotetext{*}{Only
1017: includes hours where weather was the primary cause of not taking
1018: data. There were $\sim 1300$ hours total during the observing
1019: campaign when the dome shutter was required to be closed due to bad
1020: weather.}
1021: \end{deluxetable}
1022:
1023: \begin{deluxetable}{lcccc}
1024: \tablewidth{0pt} \tablecaption{WMPol Beam Model Characteristics
1025: \label{t6}}
1026:
1027: \tablehead{\colhead{Parameter}& \colhead{Q-band} &
1028: \colhead{W-band} }
1029:
1030: \startdata
1031:
1032: DIR (dBi) & 55.65 & 62.62 \\
1033: XPD (dB) & 39.81 & 67.01 \\
1034: \% DEPOL & 0.03 & 5.6E-05 \\
1035: FWHM X (arcmin) & 19.08 & 8.45 \\
1036: FWHM Y (arcmin) & 18.61 & 8.45 \\
1037: FWHM AVE (arcmin)& 18.845 & 8.45 \\
1038: ELL & 1.03 & 1.0 \\
1039:
1040: \enddata
1041:
1042: \tablecomments{DIR is the maximum directivity of the co-polar
1043: component. XPD is the cross polar discrimination defined as the
1044: ratio between the copolar maximum and the cross polar maximum. \%
1045: DEPOL is the integrated depolarization factor defined as
1046: $(1-\frac{\sqrt{Q^2+U^2+V^2}}{I}) \times 100$ where Q, U, V, and I
1047: are the Stokes parameters. FWHM X and FWHM Y are the angular
1048: resolutions along the major and minor axis of the elliptical fit,
1049: ELL is the ellipticity of the -3 dB contour level}
1050:
1051: \end{deluxetable}
1052:
1053: \begin{deluxetable}{lrr}
1054: \tablecaption{Cuts of Raw Data \label{t7}}
1055: \tablehead{\colhead{Issue} &
1056: \colhead{Q-band (hours)} & \colhead{W-band (hours)}} \startdata Raw data & 2169 & 2169 \\\\ \tableline \\
1057: Not scanning (testing, calibration, sources) & 130 & 130 \\
1058: Short files (less than 60 s) & $<$ 1 & $<$ 1 \\
1059: Saturation (due to Sun) & 188 & 468 \\
1060: RMS cut & 173 & 295 \\
1061: Cold plate temperature & 12 & 6\\
1062: Weather image & 92 & 62\\\\ \tableline \\
1063: Total cut & 595 & 961 \\\\ \tableline \\
1064: Data remaining & 1574 & 1208 \\
1065: \enddata
1066: \end{deluxetable}
1067:
1068:
1069: \begin{deluxetable}{ccccc}
1070: \tablecaption{Comparison between recent CMB polarization
1071: experiments \label{t8}}\tablehead{\colhead{Multipole range} &
1072: \colhead{Frequency (GHz)} & \colhead{Main $EE$ result} &
1073: \colhead{Experiment} & \colhead{Site}} \startdata
1074: $66<\ell<505$ & $90$ & $<14\ \mu\mathrm{K}$ & PIQUE & NJ, USA \\
1075: & & & \cite{hedman01} & \\ \tableline
1076: $2<\ell<20$ & $26-36$ & $<10\ \mu\mathrm{K}$ & POLAR & WI, USA \\
1077: & & & \cite{keating01} & \\ \tableline
1078: $59<\ell<334$ & $90$ & $<8.4\ \mu\mathrm{K}$ & PIQUE & NJ, USA \\
1079: & & & \cite{hedman02} & \\ \tableline
1080: $30\lesssim\ell\lesssim 900$ & $26-36$ & Detection & DASI & South Pole\\
1081: & & & \cite{kovac02} & \\ \tableline
1082: $95<\ell<555$ & $26-36$ & $<32\ \mu\mathrm{K}$ & COMPASS & WI, USA \\
1083: & & & \cite{farese04} & \\ \tableline
1084: $400\lesssim \ell\lesssim 1500$ & $26-36$& Detection & CBI & Chile \\
1085: & & & \cite{readhead04} & \\ \tableline
1086: $640<\ell<1270$ & $90$ & Detection & CAPMAP & NJ, USA \\
1087: & & & \cite{barkats05} & \\ \tableline
1088: $446<\ell<779$ & $26-36$ & $<7\ \mu\mathrm{K}$ & CBI & Chile \\
1089: $930<\ell<1395$ & & $<12.8\ \mu\mathrm{K}$ & \cite{cartwright05} & \\
1090: $1539<\ell<2702$ & & $<25.1\ \mu\mathrm{K}$ & & \\
1091: \tableline
1092: $201<\ell<1000$ & $145$ & Detection & BOOMERANG & Antarctica \\
1093: & & & \cite{montroy06} & \\ \tableline
1094: $2<\ell<6$ & $23-94$ & Detection & WMAP & L2 \\
1095: & & & \cite{page07} & \\ \tableline
1096: $151<\ell<693$ & $140$ & Detection & MAXIPOL & NM, USA \\
1097: & & & \cite{wu_jh07} & \\ \tableline
1098: $170<\ell<240$ & $42$ & $<14\ \mu\mathrm{K}$ & WMPol & CA, USA \\
1099: & & & This work & \\
1100: \enddata
1101: \end{deluxetable}
1102:
1103: \clearpage
1104:
1105: \begin{figure}
1106: \plotone{f1.eps}\caption{Optical design of WMPol telescope. The
1107: primary reflector (with its parent parabola), the secondary
1108: reflector (with its parent ellipsoid), and the dewar are shown.
1109: Radiation from the sky is focused by the primary at the common
1110: focal point of the primary and secondary (the origin of the plot)
1111: and then focused by the secondary at the phase center of the
1112: central horn in the dewar.\label{od}}
1113: \end{figure}
1114:
1115: \begin{figure}
1116: \plotone{f2.eps} \caption{Drawing of the WMPol telescope with
1117: primary, secondary, dewar, and frame shown.\label{te}}
1118: \end{figure}
1119:
1120: \begin{figure}
1121: \plottwo{f3a.eps}{f3b.eps} \caption{Q-band (left) and W-band (right)
1122: filter bands. The data show transmission (normalized to input
1123: signal) through the filters. The different lines denote each of the
1124: filters used in the radiometers. \label{fil}}
1125: \end{figure}
1126:
1127: \begin{figure}
1128: \plotone{f4.eps} \caption{Schematic of WMPol receiver array.
1129: Microwave radiation enters the radiometer through the vacuum
1130: window in the dewar (right) where it is fed into corrugated scalar
1131: feed horns. \label{ds}}
1132: \end{figure}
1133:
1134: \begin{figure}
1135: \plottwo{f5a.eps}{f5b.eps} \caption{Q-band (left) and W-band
1136: (right) feed horn return loss as a function of frequency. Both
1137: modeled return loss (dashed lines) and measured performance (solid
1138: lines) are reported. \label{rl}}
1139: \end{figure}
1140:
1141: \begin{figure}
1142: \plotone{f6.eps} \caption{View inside the dewar. The W-band
1143: polarimeter horn (top, small horn) is aligned with the optical
1144: axis of the telescope and the two W-band horns underneath are for
1145: the W-band anisotropy radiometer. The Q-band polarimeter horn is
1146: tilted $4\degr$ towards the W-band polarimeter horn due to the
1147: curvature of the focal plane when off of the optical axis.
1148: \label{dp}}
1149: \end{figure}
1150:
1151: \begin{figure}
1152: \plotone{f7.eps}\caption{Contour plots of beam models for the Q-band
1153: (left) and W-band (right) polarimeters. The axes are dimensionless
1154: direction cosines given by $u=\sin\theta\cos\phi$ and
1155: $v=\sin\theta\sin\phi$, where $\theta$ and $\phi$ are spherical
1156: coordinate angles with respect to the maximum beam response.
1157: However, $u$ and $v$ are well approximated by radians for such small
1158: angles. The first three contours correspond to -3,-6, and -10 dB
1159: below the peak. Note the low level asymmetry (coma lobe) of the
1160: off-axis Q-band feed. \label{bmQP}}
1161: \end{figure}
1162:
1163: \begin{figure}
1164: \plotone{f8.eps}\caption{Determination of the Q-band beam size. A
1165: least squares fit between observations of Tau~A and a point source
1166: convolved with a Gaussian beam was used to determine the beam size.
1167: The measured FWHM beam size of the Q-band polarimeter is $24\arcmin
1168: \pm 3\arcmin$. The number of degrees of freedom is 347,
1169: corresponding to 347 HEALPix pixels (3.4 arcmin resolution, nside =
1170: 1024) that cover a 1.1 square degrees circular sky area centered on
1171: Tau~A.\label{f:beamsize_Q_chi_reduced.eps}}
1172: \end{figure}
1173:
1174:
1175: \clearpage
1176:
1177: \begin{figure}
1178: \plottwo{f9a.eps}{f9b.eps} \caption{Example of automatic
1179: calibration sequence in Q-band (left) and W-band (right), 25 Hz
1180: data rate. Placing the dielectric film at a $45\degr$ angle in
1181: front of the radiometers creates a relatively small polarization
1182: signal that is easily detected. \label{auto}}
1183: \end{figure}
1184:
1185: \begin{figure}
1186: \plotone{f10.eps} \caption{Automatic calibrator. The dielectric
1187: sheet is supported by a frame constructed of fiberglass tubing.
1188: During a calibration sequence, the dielectric sheet is moved in
1189: front of the dewar window as shown and unpolarized radiation from
1190: the ambient temperature load and the sky (through the optics) is
1191: slightly polarized after reflecting off of or passing through the
1192: sheet to the polarimeters. This small polarization signal is due
1193: to the difference in the reflection coefficients of the sheet for
1194: radiation polarized either parallel or perpendicular to the plane
1195: of incidence. After calibration, the frame is rotated away from
1196: the polarimeters about the pivot point until it rests on top of
1197: the dewar. Limit switches are used to properly position the frame
1198: and a digital flag is recorded in the data to identify when
1199: calibration sequences occur. \label{cal}}
1200: \end{figure}
1201:
1202: \begin{figure}
1203: \plotone{f11.eps} \caption{The WMPol transfer function for Q-band.
1204: The transfer function is noisy for $\ell<120$ and it was not
1205: computed for $\ell<100$. The multipole $\ell=120$ corresponds to
1206: an angular scale of about $1\fdg5$, which is slightly larger than
1207: the radius of the sky area observed. The transfer function becomes
1208: negligible for $\ell>300$, which corresponds to an angular scale
1209: of about $0\fdg6$, a value that approaches the Q-band beam size.
1210: \label{windowfunction}}
1211: \end{figure}
1212:
1213: \begin{figure}
1214: \plotone{f12.eps} \caption{WMPol upper limit of 14 $\mu\mathrm{K}$
1215: ($95\%$ confidence limit), on $E$-mode polarization, in the
1216: multipole range $170<\ell<240$. This result was obtained with 422
1217: hours of observations of a 3 $\mathrm{deg}^2$ sky area near the
1218: North Celestial Pole. The predicted angular power spectra $TT$,
1219: $TE$, and $EE$ from a $\Lambda$-CDM concordance model are also
1220: plotted for illustration. \label{upperlimit}}
1221: \end{figure}
1222:
1223: \end{document}
1224: